CVPR2025

ACE: Anti-Editing Concept Erasure in Text-to-Image Models

Zihao Wang, Yuxiang Wei, Fan Li, Renjing Pei, Hang Xu, Wangmeng Zuo

摘要

The following materials are provided in this supplementary file: • Sec. A: training algorithm and more analysis of our proposed ACE. • Sec. B: details of training and evaluation. • Sec. C: more evaluation results, including the FID evaluation, Masactrl editing evaluation, explicit editing evaluation and more ablation. • Sec. D: more qualitative results.